d c - eric · cs 001 353. miller, george a., ed. linguistic communication: perspectives for...

ED 097 622

AUTHORTITLE

INSTITUTIONSPONS AGENCY

PUB DATENOTE

AVAILABLE FROM

EDRS PRICEDESCRIPTORS

i. DOCUMENT RESUME

CS 001 353

Miller, George A., Ed.Linguistic Communication: Perspectives for Research;Report of the Study Group on Linguistic Communicationto the National Institute of Education.International Reading Association, Newark, Del.National Inst. of Education (DHEW), Washington,D C7453p.; Report of a Study Group on LinguisticCommunication in Hyannis area of MassachusettsQAugust 13-24, 1973International Reading Association, 800 BarksdaleRoad, Newark, Delaware 19711 (Stock No. 929, $2.00nonmember, $1.50 member)

MP-$0.75 HC-$3.15 PLUS POSTAGECommunication (Thought Transfer); *EducationalObjectives; *Educational Research; EducationalStrategies; *Linguistics; Reading Comprehension;Second Languages; Social Influences

ABSTRACTThis publication, which is divided into three parts,

contains the report of a study group of the National Institute ofEducation which met to investigate some of the problems of linguisticcommunication. Part 1 summarizes the general point of view of thegroup. Part 2 discusses objectives, strategies, current status, andtime scale. Part 3 describes research activities which cover thesocial and developmental context with influences outside theclassroom, characteristics of teachers and classroom, characteristicsof the reader, influence on reading of dialectal variation, and theprocesses of reading and writing, especially basic literacy,comprehension of language, writing, and second language learning. Thestudy group recommends a program of research and development onlearning and instruction in the elements of linguisticcommunication--reading, writing, listening, speaking--includinginteractions among these elements. (SW)

N.1

1/4.0

S !f. PAP T MI NT OF HEALTH.I 01.3( A .1.. hi a WELFARE

o. ,,s 'Nod NS MUTE OFt..R f ION'. t N REPRO. . vl .LO FROM

7`. I - , 4' : .4 1 niel6;15,k OPNIONS

; : %. .; . t NAM: . v RE P41.% % . T tl 06

' ''' t v

1

c)

Linguistic Communicai ion:Perspectives for Research

George A. Miller, EditorThe Rockefeller University

Report of theStudy Group on Linguistic Communicationto the National Institute of Education

Published byINTERNATIONAL READING ASSOCIATION800 Barksdale Road Newark, Delaware 19711

ii

INTERNATIONAL READING ASSOCIATION

OFFICERS19741975

President Constance M. McCullough, California State University,San Francisco, California

Vice-President Thomas C. Barrett, University of Wisconsin, Madison, Wisconsin

Vice-President Elect Walter H. MacGinitie, Teachers College, ColumbiaUniversity, New York, New York

Past President Millard H. Black, Los Angeles Unified School District,Los Angeles, California

Exmtive Director Ralph C. Staiger, International Reading Association,Newark, Delaware

DIRECTORS

Term expiring Spring 19 75

Harold L. Herber, Syracuse University, Syracuse, New YorkHelen K. Smith, University of Miami, ('oral Gables, FloridaGrace S. Walby, Child Guidance Clinic of Greater Winnipeg,

Winnipeg, Manitoba

Term expiring Spring 19 76

Ira E. Aaron, University of Georgia, Athens, GeorgiaLynette Sallie Gaines, University of South Alakma, Mobile, AlabamaTracy F. Tyler, Jr., Robbinsdale Area Schools, Robbinsdale, Minnesota

Term expiring Spring 1977

Roger Farr, Indiana University, Bloomington, IndianaGrayer A. Ransom, University of Southern California, Los Angeles, CaliforniaHarry W. Sartain, University of Pittsburgh, Pittsburgh, Pennsylvania

Library of Congress Cataloging in Publication DataStudy Group on Linguistic Communication, Hyannis, Mass.,

1973,Linguistic communication.

Report to the National Institute of Education.Bibliography: p.1. Communication Congresses. 2. Language and

languages Congresses. I. Miller, George Armitage.1920- ed. II. National Institute of Education.III. Title.Pc,7.S74 1073 410%7'2 74-13152ISBN 0-87207-929-5

iii

CONTENTS

Foreword v

Preface vii

Acknowledgements viii

I 'Introduction I

II Objectives and Strategies

A. Objectives I_'

B. Strategies lb

Current Status It)

D. Time Scale 20

III Research Activities 2I

A. The Social and Developmental Context 21

t)

I. Influences ()inside the Classroom

'. Characteristics of Teachers and Classrooms

3. Characteristics of the Reader

4. Influence on Reading of Dialectal Variation

B. The Processes of Reading and Writing 30

I. Basic Literacy

Comprehension of Language3. Writ int!

4. Second Language Learning

iv

The International Reading Association attempts,through its publications, to provide a forum for awide spectrum of Opinion on reading. This policypermits divergent viewpoints witnout assuming theendorsement of the Association.

FOREWORD

The most commonly perceived definition of literacy is the "ability to usecorrespondences of visual shapes to spoken sounds in order to decode writtenmaterials, and to translate them into oral language." The more common defini-tiohs of the word place this ability at a relatively low level of decoding and atthe lower level of the thinking process.

The authors of this publication perceive literaLy in a broader sense. Theyaver that "reading and writing are tools, not goals" and that "literacy poses aproblem for the whole society not merely for the school child and his teacher."The authors also affirm that literacy skills are acquired in many ways and inmany situations not solely within the classroom.

This publication constitutes the report by a study group organized by theNational Institute for Education to study some of the problems of linvisticcommunication. It recommends "a program of research and development onlearning and instruction in the elements of linguistic communication reading,writing, listening, speaking including interactions among these elements."

The publication by the International Reading Association of this report is aservice to its membership and the profession: it provides the vehicle for sharingwith other disciplines the ideas formulated by the study group and offerssuggested areas for needed research.

The Association expresses its appreciation to the :Nlational Institute forEducation, both for having funded this most important and productive confer-ence and for having granted permission to IRA to publish the report of the studygroup. The Association, as is the entire educational profession, is deeply indebted to the individuals who participated in the conference and who preparedthis report.

Millard II. Black, PresidentInternational Reading Association

1973-1974

VII

PREFACE

In early December 1973 the National Institute of Education established fivepriority areas. One priority area is called the Essential Skills Program. Primaryemphasis during the first year of the program has been directed toward issues inLinguistic Communication.

To a substantial degree the report of the Summer Study group in LinguisticCommunication has served to guide the efforts of the Essential Skills Program.In particular, the program has geared its activities to the two areas defined bythe Summer Study as "the most pressing national problems of literacy at thepresent time ...

1. Imparting basic literacy to those who most need it.2. Raising language comprehension in the entire population."Activities in the first problem area were already being carried out by the

NIE, although not in a cocrdinated, purposefl fashion. Roughly five milliondollars per year, for example, has been devoted during the first two years of NIEto provide continuing support to curriculum development and related readingresearch designed to improve instruction in elementary schools. Much of thiswork concentrates on the creation of new methods to impart basic literacy skillsto all children. A first task of the Essential Skills Program has been to pull thiswork together in an attempt to describe institute activities to persons interestedin reading research and development and to provide information for futuredirections of the program.

In the second problem area the Essential Skills Program has initiated anumber of the activities suggested by the Summer Study. We have attempted tocreate a balanced program in language comprehension with emphasis beingplaced bot'l on deflecting researchers in Artificial Intelligence, Visual Informa-tion Processing, and Linguistics to attend to applied issues in language compre-hension and on supporting other researchers to pursue information about thestate of current practice in teaching comprehension skills in the schools. Much ofthe work during the first six months of the program has been geared to thedevelopment of a systematic agenda for the next three to five years of NIE'sactivities in language comprehension.

Throughout this planning process we have attempted to solicit suggestionsand reactions from a wide variety of scientists and practitioners. The publicationof this report by the International Reading Association represents a major partin this effort to involve interested plies in our attempt to create as responsibleand responsive a program as possible.

Marshall S. Smith, Acting Assistant DirectorNational Institute of Education

May 1974

viii

ACKNOWLEDGEMENTS

From August 13 to 24, 173, the National Institute of Education sponsored aStudy Group on Linguistic Communication in the Hyannis afea of Massachu-setts. The purpose of the Study Group was "Tw recommend a program ofresearch and development on learning and instruction in the elements of linguis-tic communication reading, writing, listening, speaking including inter-actions among these elements." The Study Group consisted of:

George A. Miller, Chairman, The Rockefeller UniversityJohn B. Carroll, Educational Testing ServiceCourtney B. Carden, Harvard UniversityDavid Elkind, University of RochesterSamuel Gibbon. Children's Television WorkshopWilliam Hall, Vassar CollegeJulian Hochberg, Columbia UniversityWilliam Kessen, Yale UniversityHerbert A. Simon. Carnegie-Mellon UniversityMarshall S. Smith, National Institute of Education

The National Institute of Education staff consisted of John Mays, DonaldFisher, and Monte Penney. Donna Lyons served as administrative assistant.Consultants who briefed the Study Group were:

Robert L. Baker, Southwest Regional Educational LaboratoryCynthia Brown, Berkeley, Californiakohert C. Calfee, Stanford UniversityJohn Coe, Oxfordshire School AuthorityEdmund ( oleman. University of Texas at El PasoElward DeAvila, Bicultural Children's TelevisionHerbert Kohl, Berkeley, CaliforniaRichard Murphy, Educational Testing ServiceLauren ReSnick, Learning Reseacch and Development Center

Roger Shuy. Georgetown UniversityThomas Stiehl. I luman Resources Reseaidi OfficeRobert I... Thorndike. Columbia universityJane W. Torrey. Connecticut College for WomenRichard Veneiky. University of WisconsinStanley Wanat, international Reading Association

A. Wilson. National Assessment of Educational Progress

I\

On the basis of these two weeks of discussion. a first draft of this report waswritten. Subsequent editing has smoothed the style somewhat. but the mainconclusions of the report have been left to stand such as they were formulatedby the Study Group ;is of August 24. i 7?. the principal addition is Part I.which al:eniptS to summamie biiefly the general point of view at which theStudy Group aimivei.

PART I INTRODUCTION

Communication binds people together, and, of all the ways people communi-cate, human language is the most important. It is in the public interest thateveryone have the basic skills required for linguistic communication.

The most basic linguistic skill is speech. The ability to produce and under-stand spoken messages is essential for social life. Children normally acquire thisability without formal education. Educational problems arise when the spokenlanguage that the child acquires at home differs from the language used inschool.

Many social groups rely entirely on spoken language. If written language islittle known or used, people regulate their social lives by an oral tradition; evenin the United States there are communities that have developed oral subcultures.In the industrialized countries, however, many opportunities are denied to thosewho are unable to read and write. In particular, our educational system assumesthat by the fourth grade children will have acquired basic literacy skills; thosewho have not acquired them find it difficult to take advantage of subsequenteducational opportunities. Although an ability to read and write is second inimportance to the use of spoken language, illiteracy is a source of many socialand ^conomic problemh. Accordingly, a heavy emphasis is placed on reading andwriting skills in the following report.

An illiterate person must turn to others to learn of the most privatemessages written to him by friends or relatives. He must depend on others to fillout tax and social security forms; to read the fine print on insurance contracts,guarantees, leases; to transcribe his own intimate written messages. Friends andneig. `)ors must be relied on to provide transportation when public facilities areunavailable, to keep him informed about consumer options, to read to hischildren in the evening. His occupational choice is severely restricted. Depen-dence on others for help in day to day decisions and activities constantlyencroaches on his individual freedom and privacy; even when aid is substantial,

2 Linguivk Communication. Perspectives for Research

many opportunities cannot be enjoyed. Privacy and the freedom to know and todo are limited privileges for an illiterate individual.

Clearly, this nation has a responsibility to provide everyone the opportunityto overcome such limitations. This responsibility has been recognized by theinstitution of universal free education. It is not the intent of this report topropose extensive new responsibilities, but to suggest how research and development activities and applications of knowledge gained through these activities canbe used to make more effective the large resources we already devote to theachievement of literacy.

Although the United State: ranks high among nations in the literacy of itspeople, not all of our problems of literacy have been solved. Estimates of themagnitude of the problems we still face obviously depend on what level ofliteracy is taken as the indispens.ible minimum, About a quarter million youngadults are estimated to be illiterate even at the lowest level; more realisticdefinitions lead to higher estimates.

These handicapped persons come from all parts of society. In proportionalterms, however, an all too familiar picture emerges. Although the schools aremanaging to teach most children to read and write, disproportionate numbers ofthe poor, speakers of languages other than English, and members of culturalminorities do not master even the most basic skills the ability to interpretsingle sentences, or to follow simple written instructions.

When we ask about the more advanced skills and knowledge required tounderstand anything beyond the simplest written messages, the success rates ofour schools are lower than they should be for all parts of the popriation,Reading and writing are tools, not goals; they are the tools needed to change theway we think and what we think about. Highly literate people not only cansatisfy the practical demands of letter writing and recordkeeping, but also haveassess to the whole world of written knowledge. These more advanced goals arenot easy to define operationally, but they must not be ignored in settingnational policy

The most pi essing nation,d problems of literacy at the present time seem tofall under two general headings:

1. Imparting basic literacy to those who most need it.2. Raising language comprehension in the entire population.

It is unlikely that either class of problems can be solved in any singlesetting for example, in the early grades of school or by any .ingle method.Literacy poses a problem for the whole society, not merely for the school childand his teacher. Literacy skills are not acquired solely in the classroom situation.By historical accident we have inherited a conception of reading instruction thatpictures a white, middle-class school child patiently sounding out stories about

Assessment 3

Throughout this report we will speak of basic literacy, comprehenshin, and functional or practical literacy.

Basic literacy means abtlity to use correspondences of visual shapesto spoken sounds in order to decode written naterials and to translatethem into oral language.

Comprehension means ability to understand the meaning of verbal,materials.

Functional or practical literacy means ability to read (decode andcomprehend) materials needed to perform everyday and vocationaltasks.

Literacy in all of these meanings admits of important varibti, ns oflevel and content, as are described.

Dick and Jane. A recurrent theme of our discussion was the need to broadea this

picture in every direeth in toward minority coinmunities, toward adult illiter-

ates, toward the deeper proc.sses of understanding language, toward different

type's of reading materials, toward Situations outside the classroom, it is our

strong conviction that the most challenging and important social and scientific

pioblems related to linguistic skills are to he found in this area surrounuing the

traditional focus of reading instruction.

Given this brief analysis of die problems, certain general recommendations

to the National Institute of Education follow naturally. Specific recommenda-

tions and strategies for implementing them are contained in the body of the

report.

A. A SSESSME TAlthough available evidence convinces us of the general urgency of the

prohleol. we ate still not well enough informed about its exact magnitude and

prevalence in our society. Immediate steps should he taken to consider how

functional lacy should he defined and what criteria should he used as

siandaHs for its achievement. At the mie time, political and fiscal reality

demand that the NIL set feasible national goals and that the NIF, be able to

monil or success in promoting those objectiv s.

1. The NIE should establish criteria by which to measure functionalliteracy, and develop special competence in the assessment of func-tional literacy and its underlying skills. In particulai, the Nil shoulddevelop alternatives to the presently widely used grade-level criteria.

A reader's skills call he measured against those or .uhei people I normteteienced measuies) or against the skills needed. according to obstn vation to

theory, for functional literacy (criterion-ieft.srenced measures).

4 linguistic tinionunication: Perspectives for Research

1. Reading grade levels are a fot m of normreferenced measure.Criterion referenced measures could he based on:Written materials and messages used by the society at large.

3. Written materials and messages representative of theoretically basedcomponents of comprehension,

4. Written materials and messages equal in difficulty level to the listeningcomprehension level of the individual.

Most of the information we now have is of the first, normreferenced kind.Prot erly interpreted, these data may he of great aid in determining inadequaciesin nu instructional efforts, Unfortunately, however, the natur Of a normativemeasure pits scrools and groups against one another in a smseless race wheresome will always he losers. Moreover, it is only throu0 great effort thatmeasures obtained from these data can he related to the demands of the society,to the nature of the reading task itself, or to the capacities of the individual. Webelieve that grade level criteria may often he more misleading than informative.

The other three procedures for assessment ale designed to overcome thesedeficiencies. The sc..:ond approach is most relevant to obtaining basic social andeconomic parity among groups and individuals in the society. It involves thesimultaneous assessment of the critical dt. made by society on the literacyskills of individuals and the development of tools to determine individualcompetence in these skills. Great stride.: have already been made in both areas.,they should he actively pursued. Alt' gh this approach is clearly criterionrather than no' m referenced in the site) run, its reliance on the present demandsof the society implies a normative r ferent: as society changes. so presumablywill the demands on the literacy skill!, of individuals.

The third and fourth approaches should be less subject to the winds ofchange. Both approaches, however, will require considerable research and timebefore they can he implemented. The third assumes the existence of a deepertheoretical understanding than we at present possess. The fourth defines a

criterion that takes individual differences into account, but too little is presentlyknown about the reliability and validity of different ways of measuring equalcompetence in spoken and written communication.

In addition to the technical advantages and disadvantages of each of thefour approaches, their social and political implications need to be weighedcarefully.

B. FIRST TASKS

Although at present we lack sufficient information to define accurately theprecise scope of the problem, our initial assessment indicates that we should notwait for further research to start or lumpily programs at both the basic andfunctional levels for those groups or individuals most in need, since it is clear

Culture and Motivation

that many people have not attained an adequate level of functional literacy

under our present system of instruction.

II. The NIE should supprrt research and development designed to findmeans of raising literacy to the levels required for effective functioning

in society.

Although functional literacy remains to be fully defined, we are at a point where

we can identify a set of face-valid, functional, reading tasks.

Ill. The NIE should give first priority to research and developmentaimed at improving the acquisition of basic literacy by persons fromlow income families and minority communities.

As stated, recommendations II and III are less precise than we wish they

were. The body of the report suggests possible criteria and strategies for the

definition of goals in both areas, basic and functional; we did not consider it our

prerogative to choose among the alternatives. We strongly suggest that the NIE

program in linguistic communication take as an initial responsibility the task of

defining goals for these two foci. Whatever specific goals are defined, however,

our analysis suggests that there is available a set of general approaches that will

facilitate their accomplishment.

C. CULTURE AND MOTIVATIONHow much a student will learn depends on how and how much he studies;

these, in turn, depend on his motivation to learn. Better ways of sustaining astudent's motivation to learn to read and write are needed.

Analyses of methods for imparting communication skills must take account

of the particular needs of the concerned groups. The audience fol such efforts is

quite heterogeneous; no single curriculum will serve them all equally well. Even

those whose first language is Spanish come from quite different cultural back-

grounds Cuban, Puerto Rican, Mexican and this situation is different from

native Americans or those who speak the black English vernacular. Cutting

across these cultural differences are the different requirements of illiterate adults

who have little motivation to read the adventures of Dick and Jane. There hasbeen considerable attention paid in recent years to the advantages of individual-

ized instruction for each child; comparable attention must be paid to theindividualization of instruction for each culture. Unfortunately, many basic

questions remain unanswered: Is there convincing evidence that black children

will learn more rapidly it' they begin reading the black vernacular? Should

children from Spanish speaking homes learn to read Spanish first? Can teachers

he prepared to develop reading materials of their own in such situations? And so

on. These and related problems are discussed in the body of the report.

6 Linguistic Communication: Perspectives for Research

Since reading instruction is a major task of the early grades, the match ofschool to home environment may have particular relevance to the developmentof linguistic communication skills, We do not feel that it is enough .simply tomake reading materials reflect more accurately an indiviatial's cultural heritagealthough that is an important start. Other aspects of schooling may have to betailored to a student's cultural background. We know that the social context isimportant socioeconomic status of the family has been one of the bestpredictors of a child's eventual success in learning to read and write but weknow little about how the social context affects learning. Some now believe thatspecific environmental factors can be identified, so '!tat the real problems can bemore clearly defined.

Efforts to integrate the various linguistic experiences of the child need to bepursued further. In order to sustain the student's motivation to learn, readingand writing should be integrated with other classroom activities, should be madeto fit the interests and purposes of the individual, and should be carried out inthe home and other places as well as the school.

IV. The NIE shoulda, explore ways of relating the curriculum and the school to thechild's home and peer culture, andb, actively promote the development of strategies for raisingliteracy by integrating reading and writing into the life and inter-ests of the individual, outside as well as inside the classroom.

We believe that this recommendation represents the best general approach tomaintaining the student's motivation, and that it can be implemented both forthe teaching of reading and the teaching of writing.

Reading conceived as a separate course of study "Put away your comicbooks, children, it's time for Reading" reading confined solely to basalreaders, reading taught as a highly stylized and rule-bound task, all tend to pushthe idea of reading as communication, as power and discovery through com-munication, into the background. Reading must he made interesting and rele-vant. A child intrigued by at mechanics might be encouraged to bring manualsto class, and to bring life to the manuals by working on engines within theschool,

As for writing. an initial deemphasis of certain stylistic rules, letter-perfectspelling and punctuation-perfect paragraphs, with more emphasis on writing ascommunication (without distracting corrections) should create an atmosphere inwhich communication is fun, in wnich both reading and writing (especially,reading what the student himself has written) take on additional interest.

Children should not he led to think of reading and writing as tasks unique tothe schoolroom. It' literacy is the responsibility of the whole culture, its advan-tages should he seen outside the schools, Alternative opportunities for develop-ing reading skills are needed for adult illiterates in any case. And reading

How People Communicate 7

instruction for children in the classroom should be supported by a world outsidethat will contribute to their general knowledge base.. Special attention should begiven to possibilities for reading instruction in the home and via the mass media,

particularly television.

D. UNDERSTANDING HOW PEOPLE COMMUNICATEThe recommendations above are based more on general experience than on

an expl:cit theory about how people communicate. Much further exploration ofthe theoretical aspects of reading acquisition and comprehc., .)n is needed.Although the process is complex, understanding the ability to decode, interpret,and comprehend written materials has advanced significantly in recent years,particularly through the work in cognitive psychology and a.. tificialIn order to build a scientiti.: base for future application,, Ills work deservesattention and support.

V. The ME should actively support efforts to understanda. the cognitive processes involved in acquiring basic readingskills andh. the cognitive processes involved in comprehending linguisticmessages.

Comprehension is the purpose of reading, yet we know far too little about theknowledge and conceptual organization needed for advanced reading com-petence. Although we have learned much about the legibility of type, patterns ofeye movements. rates of information processing, and the like, these facts havenot been put together in a coherent scientific theory of reading. We mustunderstand better the higher menial processes that control the intentional act ofreading. Because principled improvements in current practice will depend onsound theiny the skills to be taught. recommendation V deserves highpriority.

IMPLEMENTATIONThe NIE must employ personnel competent to direct and supervise efforts

to implement these recommendations.

VI. The NIE should employ staff and make administrative arrange.ments to undertake effective programs of research and development onthe educational problems of linguistic communication.

All other recommendations presuppose the existence of an effective unit of this

kind.

8 Linguistic Communic.:;an: Perspectives for Research

F. A FINAL THOUGHTA recurrent lament heard durig our discussions was that professionals

concerned with linguistic communication do not communicate with each other.Teachers of reading have no easy way to learn of effective instructional tacticsdeveloped by other teachers. Research workers interested in the reading processcome from such a variety of professional and academic backgrounds that theyare often unaware of their common objectives. And the greatest gap is betweenthe teacher and the research worker.

Our own attempts to understand this failure of communication led us tospeak of "the predicament of perspectives ;" teacher, researcher, and studentoften have very different conceptions of reading and of learning to read.Teachers shottid adopt the widest possible framework. Their concerns, eventu-ating in procedures that will be effective in the classroom, are with the detectionand diagnosis of potential reading problems, selution of effective methods ofdealing with them, and instruction itself. Proper detection and diagnosis requirean understanding of both the scientific perspective an effort to analyzecomplex skills into their component processes and the child's interest inreading and writing in relation to the organic growth of the child's general skillsin linguistic communication. The transition from a knowledge of what theproblem is to what to do about it is most often left undetermined by thescientist. The materials of instruction should be related both to theory and tothe c'tild's needs.

VII. The NIE should support efforts to make explicit not only thescientist's theories of reading and of learning to read, but also theperspectives of the teacher and the student, with the aim of facilitatingcommunication among them.

This recommendation assumes that the critical step in improving comtnunicationbetween scientists and teachers, and between teachers and students, is to makeexplicit the differences in bask assumptions with which they approach theircommon concerns.

9

PART II OBJECTIVES AND STRATEGIES

The nation has often reaffirmed its responsibility for providing opportunity to

all citizens to become full participants in the national life as producers, asconsumers. as citizens, as human beings. To the extent that understanding thespoken and written word is important for exercising that opportunity, it be-comes a national responsiblity to see that all citizens are literate.

However, we do not wish to exaggerate the importance of comprehension

skills, and especially reading skills, in the routine of everyday living. The reading

requirements of many occupations are quite modest; probably only a relatively

small proportion of all jobs require more than current eighth-grade readingcomprehension.* Moreover, the formal admissions requirements for occupations

sow -times overstate the real importance of reading skill, or of other formal

training, to job performance.A persistent concern of members of minority communities, most vocally

expressed by black Americans, has been their record in passing standardizedemployment tests. One reason given for this state of affairs centers on themismatch between most standardized employment tests and the actual require-ments for jobs. Indeed, some recent data suggest that many tests for jobs such asthat of patrolman might well he used to admit students to the most selective of

our colleges. Readability estimates place much of this material well beyond theformal educational level required to qualify tor the job. The extent of literacy in

fact required for employment can only be ascertained through job-analyses and

other studies of a carefully constructed sample of a variety of occupations.

Even the most bookish among us depend on high-level reading skills only

during a ,,,.,III portion of each day. Nor is there evidence that raising aneighth-grade reader's skill to the twelfth-grade level will contribute much to hisperformance as a voter. Voting in a democracy rests on no presumption of

*We would have preferred to express this point by citing criterion-referenced test results.However. until adequate criterion- referenced test% are developed and utilized we arcobliged to depend upon norm-referenced measures.

10 Linguistic ammunication: Perspectives for Research

universal expertness on subjects of public policy: rati.er, it rests on the observa-tion that adults of ordinary intelligence and education can he competent judgesof the integrity and performance of public officials. and shrewd judges, too, ofhow to protect themselves from injurious abuses of public power.

Having stated all these reservations, we must nevertheless conclude that aperson with limited reading skills may he deprived of many benefits of oursociety. The importance of literacy cannot be measured by the number ofoccasions many of them quite brief when being able to read or not opens orcloses a door of opportunity, shields one from an accident or exposes him to it,provides one with a sense of competence and independence or makes 4.nedependent on others.

The skill of reading provides the reader with access to help-wanted ads inthe newspapers. and to announcements of store sales. It allows him to find hisway about the streets, (A. visitor to Tokyo can experience something of theilliterate's problems of navigation.) It permits him to read instructions on a

pesticide bottle and recipes in a cookbook. It permits him to read his ',tail, fillout his Social Security forms, obtain a driver's license.

'the public responsibility to provide opportunities to learn to read and writeis discharged primarily by the American public school system which offersuinver:.1 free education through the twelfth grade. Thus, national objectivesmust he translated into objectives jOr the educational syste n since the schoolsare the most important instruments available for achieving our objectives.Although there is a sizeable number of adult illiterates who should be offerededucational opportunities, our prevailing national strategy is preventive toimprove tl.e teaching of reading and writing in order to prevent more childrenfrom growing into illiterate adults.

Perhaps because universal literacy has alv ays been a goal of our educatilnalsystem. familiarity has led us to assume we know what it means. Programsadvocating every person's right to learn to read receive broad public attention.Yet these goals or rights are inherently vague. What level ; t0 competence is to becounted a; literacy? Do we mean literacy in Fnglish or in any language? Whatparticular language skills are involve0 What degree of accomplishment in eachshould he described as the indispensable minimum? When one attempts to assessthe current situation. the range of language skills achieved by different individ-uals becomes apparent. If one defines the national objective at a relatively lowlevel, the magnitude of the problem can he underestimated. The higher thenational level one aspires to, the greater the number who will need educationand the more difficult the achievement will become.

Somewheie. ilear the core of the responsibility accepted by our schools is apicture, shaped b tradition. of a white. middle-class child sitting quietly in aclassroom patiently sounding OM the words in simple stories. When tine focuseson this picture. o seems reasonably pleas.mt that is not where our national

Objectives and tittategies I I

profile:' literacy originate. As one expands this view to include minoritygroups, illiterates, understanding what is sounded out, alternative readingmaterials, md the out-of-school uses to which reading ability is expected totransfer, the problems begin to appear.

Language minorities face special problems, not just in reading, but inmaking full use of the educational opportunities available to them. In the UnitedStates mos, of the information we want citizens to have is communicated instandard Fnglish. This is true in the mass media geieral.y, as well as in theschools. lhe student's knowledge Of English is thi.s the cornerstone of oureducational plan. Some of the most serious problems confronting our educa-tional system today arise from the fact that students particularly students inthe inner cit IL,: do not enter the schools with equal linguistic competencies.Some speak no English at all, some speak nonstandard English. and even thosewho speak English will not have mastered all the language skills they will requirein the course of their educational experience. In this connection, William Labov11972) has written the following.

Many skills have to he acquired before we can say that a person haslearned standard Lnglish. The following list is a scale of priorities that Iwould suggest as helpful in concentrating our attention on the mostimportant problems:

a. Ability to understand spoken English (of the teacher)h. .Ability to read and comprehendc. Ability to communicate (to the teacher) in spoken Englishd. Ability to communicate in writinge. Ability to write in standard English grammarf. Ability to spell correctlyg. Ability to use standard 1.nglish grammar in speakingh. Ability to speak with a prestige pattern of pronunciation

(and ,boid stigmatt/ed forms). ( p. 5)

We usuall assume a and take h to he the firs: academic task for the child. Butwhen a is deficient , its amelioration should to DI iority over b,

I bov's li.t .me. to remind us of the cmolexity of the task faced by theschools. ind of ihe corresponding co:nplxily that must he incorporat , into

welldetined national objective. And even if we had a well-defined objectiveand a valid assessment of out curiet distance from it. we would still facedecisions as to which act.ons would most effectively reduce the distance. Thehulk of the present report is concerned with this third, tactical stage of thenational progi am.

De !tuition of litoac have been as wide as they have been deep; at onetime of another. oral speaking and writimu skills have been related to definitionsof Iner.ic. We must evimine sonic of the existing definitions and objectives, notonly lot their theoretical and functional relevance, but also in the light ofpolitical practicality.

12 Linguistic Cmmunication: Perspectives for Research

A, OBJECTIVES

1. Basic Literacy

The discussions of the Study Group used a rough distinction betweena) bask or initial reading skills, and b) comprehension, particularly as it involvedfunctional, practical, or advanced reading skills between the skills needed totransfin.n: written symbols into speech and the skills and knowledge needed inoccupations, and on consumers and citizens, We recognize the artificiality of thisdistinction, since at every level of skill reading is an effort after meaning, yet itserves as a convenient expository device for separating out the rather differenteducational problems involved in each case.

The distinction can also serve a useful purpose in formulating objectives.The skill of correctly decoding written language could serve as a standard ofbasic literacy for our society. One definition of "basic literacy" might he theability to read aloud a selected list of words; a second might be the ability toread aloud selected sentences.

Extent of problem. We believe that, with few exceptions, everyone couldachieve basic literacy with proper instruction and support. A very large fractionof the population has progressed well beyond this minimal criterion, yet thereare pockets of people for whom basic literacy is yet to he achieved.

We know of no attempt to detect and determine systematically the degreeof failure to achieve this minimal level in the general population. Such datacould have great importance for national policy. Unless there are reliable data ofwhich we are unaware, we recommend an assessment to determine specifically towhat extent an objective of "universal minimal literacy" has been accomplishedin the adult population. The survey would he a 1,:obability sample, weighingthose subgroups in the population where an inability to read words and sen-tences aloud might be expected.

I Comprehension and Functional Literacy: OccupationWith improvements in our methods of measuring literacy and the cognitive

pros, sses associated with it, we may soon be able to define a national goal invich a way that we could adapt it to the wide var4-(y of individual circumstancesand, at the very least, be able to recognize when we had achieved it. One line ofattack would he to ask the educational system to give e-eryone the ability toread up to the level required of him by his occupation.

Although there are clear virtues to such an approach, this line of thinkingdoes not seem to lead to a satisfactory definition of a national goal. Not only aremost of the reading students in the schools too yo.ing to have any occupation.but it would he difficult to avoid the inference that no e had a right to readbeyond the requirements of his occupation that the ambitious worker wouldnot have educational support in attaining jobs beyond his present level ofreading ability.

Objectives 13

As an alternative, we might want the educational system to bring everyoneto that reading achievement level which at least 80 percent of those gainfullyemployed were required to have. This would lead to an extensive survey of theliteracy requirements for all definable occupations, and to the deve! 'pment ofinstructional materials specifically aimed at each.

Extent of problem. Preliminary evidence suggests two key points:

a. Scattered data indicate that only a relatively low level (8th grade) ofcommunication skill is sufficient to acquire the knowledge to perform ade-quately in many occupations.

b. Other data indicate that there is little relation between reading compre-hension and on the job performance ratings within occupational categories.Taken together these data imply that most people could achieve the communica-tion skills required to perform competently in most occupations.

Were these data substantiated, they might also alert teachers. parents, andworkers to the possibility that occupational opportunities are far more availablethan previously imagined. Too, it would give teachers some absolute perfor-mance criteria to aim at criteria related to reading skills demanded by thesociety. And if employers and unions realized that only a basic level of literacy isrequired to perform adequately in most occupations, the entrance requirementsto many jobs might be reduced. Similarly, other literacy demands made by the

society might be eased when, for example, lawyers and advertisers realize thatthey are not communicLting with some members of society. At the very least.these data would focus attention on how to change societal activities that arecurrently selective and exclusive.

We recommend that two types of data be gathered for a random sample ofoccupations in the society;

a. Data on the level of reading skills required to have access to theoccupation.

h. Data on the level of reading skills necessary to gain the knowledge to beable to perform adequately in the occupation.

3. Comprehension and Functional Literacy: Consumer ChoiceAnother approach to specifying functional literacy is suggested by placing a

high value on the notion of consumer choice. Consumer choice implies that mostpeople have the skills and knowledge to make informed decisions when faced

with impot ant life choices.A decision to maximize the potential chokes open to an individual could

involve us in a wide variety ot activities. A basic core of skills and knowledge can

he enumerated it there are universal situations situations that everyone can bereasonably expected to encounter where reading and thinking skills are re-quired to make informed decisicas. Activities dependent on reading and thinkingability and related to the areas of need immediately present themselves.


a. Consumer Economics: product labels, advertising, cookbooks, nutri-tional documents, shopping ads, yellow pages, consumer reports.

b. Health: life insurance, health insurance, health instructions from schoolto home.

c. Transportation: schedules, place names, directional signs.d. Government and Law: voting literature, apartment leasing contracts,

rights as a citizen.

e. Community Resources: telephone directory, appropriate newspaperarticles, bulletin board notices.

f. Occupational Opportunities: want ads, employment brochures.Skills needed to comprehend materials in these areas should be understoci

and appropriate assessment instruments developed and utilized. This woulddisclose pockets in the population where people do not have the skills andinformation to exercise real freedom of choice. Resources might then bedirected toward alleviating deficiencies in knowledge and basic skills.

Extent of the problem. In order to pursue this approach, we offer a varietyof recommendations. At present, there are at least three on-going efforts toassess functional literacy in the nation. The Educational Testing Service is justcompleting a national survey of reading skills the items on the survey weresuggested by prior survey of the reading habits and requirements in the popula-tion. The National Assessment of Educational Progress will be carrying out asimilar survey in the very near future. Northcutt (1973) has also carried out asimilar assessment. We recommend that NW pay close attention to the results ofthese surveys and engage in active communication with the three grotips aboutUnplications, problems and gaps in the data. This activity should he initiatedimmediately.

4. Equalization of Subgroup Norms

One objective advanced has been the attainment of similar scores or achieve-ment levels on reading tests by various population subgroups. Literacy is, underthat objective, defined as that which is measured by the standardized readingtests, at a level for all population subgroups equal to the grand mean. Until thenation takes dramatic steps to -equalize" the quality of life in general acrosspopulation subg.oups. we will continue to find group differences in literacy.Such group differences might disappear or need radical reinterpretation were weinstead to develop criteria for literacy based on theoretical and functionalnotions that can he used to guide local and national policy.

Furthermore, to define our goals in terms of norms for the eighth or thetwelfth (or any other) grade can he misleading, for the norm provides a movinginstead of a stationary target at which to aim. lt' the average reading skills of thewhole high-school population were raised substantially , but we used norm-referenced tests to measure literacy and constantly reset the norm, it would

Objectives 15

appear that no progress had been made. The average student would still beaverage, and half the students would still fall below the norm, as before, Progresstoward literacy goals must be measured against standards based on theoretical OrJunctional criteria, rather than norm - referenced standards. We must measure thereading comprehension levels that are required to handle specific kinds ofreading materials that are used on the job, in the home, and in other lifesituations, and measure progress against these stable markers rather than theribber riders of norm - referenced standards:

One strong evidence that levels of literacy and language understanding arenot fixed once and for all is the significant trend in average absolute scores onthe same measures that have taken place over the past fifty years (see, forexample, Tuddenhatn, 1948; Schrader, 1%8; Jencks et al, 1972, p. 63). A roughextrapolation of these data over the past fifty years leads us to estimate that wecan expect an increase in average test performance of between one sixth and onequartet of a standard deviation per decade. Naturally, this rate of increase is notinevitable in the future, nor may we attribute it entirely to changes in schooling.

In order to evaluate progress toward our national objectives with respect toliteracy, such studies should be replicated preferably with measuring instru-ments based on criteria that do not shift over time, and with reading achieve-ment, rather than general intelligence, as the variable to be measured. Thequestions that must he answered before this can he done are discussed inPart III.

5. Individual PotentialAnother line of attack on the formulation of a national objective is to ask

the educational system to provide the opportunity for everyone to read up tothe limits of his own ability. The obvious objection to this approach is that itseems to require us to measure the limits of Ln individual's ability to read, and todo so at a time when our methods for measuring how well he can read not tomention how well he mold read still need considerable improvement. There isa way to avoid this objection, however, that seems natural and well within thelimits imposed by the present quality of our reading yadsticks. Suppose wewere to consider two measures: a measure of the person's ability to comprehendspoken language, and a measure of his ability to comprehend written language.1 hen the task would he to devise measuring instruments that would enable us todetermine for any given individual whether his reading comprehension wascommensurate with his listening comprehension of English messages. Givenappropriate instruments, a well-defined national goal would he to educate everyreader up n a level commensurate with his listening comprehension.

Extent of problem. The development and validation of a testing procedurethat would provide such a reading jilotient might he a major undertaking. Notonly might considerable research on listening comprehension and its relation to


reading comprehension be required, but special adaptations of the instrumertwould be required to deal with bilingual and bidialectal individuals. On the basisof the discussions we have been able to devote to this topic, the Study Grouphesitates to recommend the development of such an instrument at the presenttime. We do recommend, however, that a further group be assembled to pursuethe possibility in greater detail both with respect to the social and politicalconsequences of formulating a national goal for reuciing in these terms, and withrespect to the various research and development strategies that would mostlikely lead to a valid and reliable method of testing. We recognize the possiblitythat this approach may turn out to be more appropriate to individual diagnosisthan to national or group assessment this recommendation, therefore, appliesequally to the diagnosis category (see page 28).

B. STRATEGIESWe must talk in the plural of the "problems of literacy." No single line can

be drawn between those who are illiterate and those who are not. A person whocan read the sports page may be tillable to read the financial page (and viceversa). Or a person who can fill out a job application blank may be unable toread the manual that describes how to use a piece of equipment needed on thejob. But these are relatively advanced levels of reading. If we consider thebeginning reader usually a young child entering the first grade his first steptoward reading and writing is to learn how to connect the printed word and thespoken language with which he is already familiar. Because it is so important,making this connection is sometimes regarded as the only problem of reading.Unfortunately, the real problems are much more complex.

I. Basic Literacy

Understanding the idea and purpose of representing familiar language and itsmeanings in printed forms, and acquiring a good level of skill interpreting suchrepresentations, are the central. goals of the first two or three years of readingstudy. Crucial to the whole enterprise of learning to read is the child's attitudetoward the task. A positive attitude is more important than acquisition of theability to recognize a few letters or words. If these skills are acquired at the costof the child's sense that reading is exciting and interesting, the price may hemuch too high. Positive attitudes towards reading seem best inculcated by anenvironment rich in reading materials, by adults who read themselves and whoread regularly to children, and by the opportunity to express and representexperiences verbally and in print, and by entertainment derived from reading.

For the child embarked on such a learning experierce, learning to read is anemotional and cognitive development that must form an organic whole, anintegrated system of closely related skills that is never completely perfected. Forthe scientist and the teacher, whose tasks are to understand and supervise this

Strait gies 17

organic growth of reading skills, a more analytic conception is required. For thechild, reading and writing are a omtiuation and extenshot of oral languageskills: .tbr those who would hope to aid the child, reading and writing areowtple.v packages of ec,ntponent skills that must be unfolded to the child insome sequence intelligible to him, For the child, reading and writing represent anunusual fvusing of :mention on the linguistic aspects of communication, withprogressively less support from the situational and linguistic context on which hehas relied to acquire his oral mastery of language; for the scientist and teacher,this focussing is merely one of many abstractions that characterize the analysisof linguistic communication via the written word.

The scientist and the practitioner, therefore, find it convenient to drawdistinctions that can organize their research and teaching, but that are merelydifferent aspects of what the child experiences as an integrated whole. Thus, theanalyst will frequently distinguish between certain basic skills required foridentifying words and phrases from print, and certain advanced skills requiredtau compiehending and using the information that the written signal encodes.Rut to the child, these are related aspects of his experience. lie does not firstlearn to identify words from print and only then begin to think what they mean,From the beginning the child's task is to recover meaning from the writtensymbols; what chinges is not his goal, but the nature of the obstacles that standbetween him and his goal. At first his worst problems seem to be with identifica-tion and recognition'. later his problems become those of comprehension andinterpretation,

It a person's search for meaning is impeded by his inability to recognize thewords represented by the written symbols, he presents problems of basicliterac,... it his s.:arch for meaning is impeded by his inability to understand themeaning of the words and sentences he has recognized, he presents problems ofcomprehension. The kinds of help required in these two situation: seem suffi-ciently different to justify the distinction as long as one does not conclude thatcomprehension is irrelevant 'luring the learning of word identification, or thatword identification is irrelevant during the learning of interpretive skills.

the ability to sound out the words that the text represents, regardless ofwhether they are understood loosely speaking. the ability to decode is the

easiest of the reading-related abilities to assess. Even this apparently simple skilldepends on many components: on being able to recognize the shapes andarrangements of letters and let toi-groups; on being able to move one's eyes withappropriate precision and control: on being able to produce acceptable speech

sounds in response to the visual patterns that comprise the printed text.

these skills are often not easy to acquire. We may he able to make a taskmore purposeful by various insttuctional innovanims, but we cannot make ittrivial. It conditions for silstallINI attention and motivation are not inaintined,in appropriate emotional climates. the decoding ability may not he aco:I're.1.

18 I.inguistic Communication: Perspectives pi. Reward'

Moreover, many young children appear to lack some of the componentskills that contribute to the decoding ability, skills that they would normallyacquire later in the course of their perceptual and cognitive development. If wecan determine specifically which skills are missing, we can at best compensatefor their absence, or at worst avoid drawing on abilities which the child does notyet have. Rather than gross assertions about "reading readiness" or "perceptualinadequacy," we need ways of determining what specific abilities are absent, andwhat roles they play in We development of literacy.

The number of persons facing problems of basic literacy is quite small in theUnited States perhaps as low as half of one percent of the population ofyoung adults. However, there is a signitican; additional number of persons whoseinitial reading skills are marginal and inadequate as a foundltion for building.

Comprehension Skills

The factors that determine the reading ability of a ten-year-old, a high-school student. or an adult are seldom the same as those that limit theperformance of the beginning reader: their reading problems have become theproblems of threading their ways through complex grammatical forms and oforganizing the ideas they extract from reading into coherent structures that canhe retained and related to their experience and knowledge.

The problem of literacy at this level of comprehension may not be specifi-cally a reading problem at all. It may be a much more general problem of abilityto understand language, whether written or oral. Same evidence suggests that if aperson with initial literacy can understand a spoken message, he can understandthe identical message when it is presented in writing. If further research substan-tiates this finding, it has two important implications for education:

. a. To the degree that read* comprehension difficulties depend on gen-eral language understanding, educatfimal programs will be siuvessful only if tin it'are effectipe in raising such language understanding.

b. the most Ofeetire instruction pr raising levels of complehension maynot be reading instruction at all.

What may he needed, for example. is broader real-world experience with thevarious kinds of objects and events that books and articles describe. Or. it mayhe chat instruction in "how to think" or "how to solve problems" wuuld be, forsome, the most efficient route to functional literacy. Or, perhaps skill at

cimiprehension at inure advanced levels better practiced and learned directlyin the context of specific content, as a byproduct, say, of learning history orautomobile mechanics o mathematics.

That we do not know the answers to the questions posed here should notprevent us from making vigorous efforts to give everyone the ability to read theworld of print that surrounds and permeates his everyday life. However, deter-mining how many people do not possess adequate skills in comprehension is

Current Status 19

inure difficult than detecting weaknesses in basic literacy, for we have fewmeasures or the levels of reading ability required to cope with the writtenmaterials encountered in various situations of everyday life,

All of the available statistics at the present time indicate that the problemsof literacy are most acute for those who are poor or culturally different frommajority groups. Is this a literacy problem? Or is it a social and moral problem?We are inclined to regard illiteracy as a symptom of deeper national problems,The real challenge, therelOre, derives less frfmt the' numbers of persons who arehandicapped socially by inadequate' linguistic skills than from the democraticobligation to provide equality of oppewunity to all citizens,

C. CURRENT SlATUSThe teaching of reading is heavily institutionalized in the United States. A

few children learn to read before entering school; a few school children maylearn something about reading from television; but the overwhelming majority ofAmericans are taught to lead in the first grades of school. This is not to say, ofcourse, that reading competence is independent of events outside the classroom,The persistent relation found between chara:teristics of the home (indexed bySl'.S of quantity of resources) and ability to read suggest that the child is, insonic way, piepared for the tasks of reading by what happens to him at homebefore school. The dimensi.,ns of that preparation are by no means clear but willmost cei wilily range from genetic influence to specific tuition in reading skills,

We are even less well able to measure the contribution to reading espe-

cially, to comprehension of the child's activities alone, on the playground, inthe streets. It' one takes the position that general understanding of his experienceis a critical factor in the child's (or the adult's) ability to use language, then thevariety of the child's out-of-school, ; :at -of-home experience and his oz-,mizationof that experience become relevant to his ability to comprehend what he reads.

1 lie first formal encounter with reading in !h..' first grade can be charac-terized in general terms. The child is one of about thirty who share flit, sameclassroom teacher: he is exposed to basal readers written in a limited vocabulary

about episodes imagint to be of interest to all children, but the amount of timelie actually spends discovering the con espondence betw.,en the written and thespoken language is usually slight: his initial skills are acquired in a sequence thatdepends on the teacher's strategy he learns phonic rules, or is taught torect)grize words printed on flash catds, or (less often) is coached to pronouncesyllables, or the hke: and he is expected to achieve a predetermined first-gradelevel of reading skill by the end of the school year. Sonic of the disadvantages ofthis instructional experience are obvious, and sonic children are fortunateenough to escape it in schools where strict grade levels are not maintained, whereinstruction is individualized, where attention is paid to the child's interests in


reading, where the child can advance at his own rate but these are exceptionsto the usual rule. Despite its limitations, however, the usual instructional systemseems to work at least for children whose prereading and out-of-schoolexperience has prepared them for it.

When we ask what is being done for those who have passed through thisinstructional system without acquiring basic literacy, the answer seems to be,"Very little." When we ask what research is receiving support, most of it seemsto be directed toward the established practices of the traditional system ofreading instruction. When we ask what is being done to match reading instruc-tion to the cultural background of the student, we find many small, interestingexperiments in progress, but no national program. When we ask what is beingdone to teach reading-as-understanding, no one seems to MOW how it should bedone.

In short, we are presently meeting our responsibility to develop initial skillsin linguistic communication for the majority, but extensions of the existingsystem to deal with minority problems are being neglected, and potentialimprovements in the existing system are often resisted when they conflict withestablished policies.

D. TIME SCALE

The Study Group avoided imposing a linear sequence upon the tasksrecommended throughout the report. Some long-range research tasks are soimportant that they should he begun now even though they may not becompleted for many years. Other tasks can he scheduled for short or mid-rangecompletion, depending upon their importance, visibility, and feasibility. Inthinking about the time scale it is important to remember that it takes someseven to ten years from the initiation of a major curriculum program until itslong-scale introduction int; the school system. Although this time-lag might hesomewhat reduced, at present it must he regarded as one the facts of life ineducational research and development. Thus, although certain actions should heundertaken immediately, their impact on the educational system may not berealized until far in the future.

21

PART Dl RESEARCH ACTIVITIES

A, THE SOCIAL AND DEVELOPMENTAL CONTEXTEven in a limited discussion of research on the problems of literacy it is

necessary to move beyond considering the child as a social statistic. Effectiveprograms must view the child as a developing individual, influenced by morethan his school environment, actively trying to extend his skills in linguistic

co In this section we look at some of the lines of research thatmight profitably he undertaken to increase our understanding of the social anddevelopmental context of learning to read.

Although research in this quadrant of the field of reading is voluminous, theresults have attained neither the precision of method nor the sophistication ofanalysis that characterizes current research on the reading process itself. How-

ever, it is our opinion that accurate diagnosis of the sources of illiteracy will haveto draw heavily on considerations of human development and interaction. Thefollowing pages will treat, in order, influences outside the conventional class-

room, characteristics of teachers and claroonic. eharacteristics of the reader,assessment and diagnosis, and the influence on reading of variations in dialect.

I. Influences Outside the ClassroomOne of the rare points of general agreement in the lore and literature of

reading holds that the child's success in learning to read is tied essentially to hiscomprehension of spoken language. A more arguable form of the proposition isthat the learner's effectiveness in reading is most intimately dependent on thecomplexity and fluency of his conceptual or cognitive structures structuresthat may he even more fundamental than speech. Whichever form of theargument survives research, the child's preparation Jr reading takes place beforeschool and outside the classroom. When we speak of basic literacy, we arctalking about making connections between patterns of writing and some aspectof what the child already knows and what he can already say When we speak of


comprehension, we are talking about the child's scores on tests, but also abouthis ability to use in his life beyond the classroom what he sees in print. And,beyond the preparation for reading that makes the child conceptually ready toread before he looks at his first pre-primer, there are a complicated and poorlyunderstood set of motivational and attentional factors that influence his induc-tion into literacy. These, too, are set underway during the years of his life beforeschool.

"Preparation for reading" is accurate enough as a first statement of theinfluence of home, media, peers. and the street on learning to read, but thephrase may contain a misleading message. Some children are prepared very wellfor the kind of reading instruction they will receive: they possess, at age six,conceptual and motivational structures that fit the first-grade classroom theyenter. Other children (and they have been disproportionately the poor, theblack, the speakers of languages °die than English) will not be prepared for theclassroom they enter. It is not proper to conclude that their preparation isdeficient or that the remedy lies in changing (or, less troublesome, merelyblaming) the "ill-prepared" child and his family: rather, we need to recognizethat reading, particularly the kind of reading jbr use that we would emphasize, isprepared for in many different ways in American life. and that schools must headapted to that diversity of preparation. We should he past the time when theintellectual culture of the first grader is markedly narrower than the intellectualculture he knows outside the school. We should he, but we are not. In theproposals for research that follow, a guiding principle has been the need betterto connect what the child brings to his first grade and what the first gradepresents to him.

a. The preparation of readers and the match to school, There are sizabledifferences in early reading ability existing among children of different socio-economic and ethnic groups. Research of the last decade has begun to studymore carefully the characteristics of families that prepare children for present-,lay school procedures. Middle-class children, for example, are said to Ime 115.n"taught to formulate and test hypotheses, to identify solution:: tentatively, toexpect and to deal with uncertainty, to respond to verbal rewards" allcharacteristics that may influence their learning to read. Research in explorationof such variation should he continued with two important modifications. Firstand fiiremos:, research on variations among different groups should not headdressed exclusively to the test of success in learning to read in present-dayconventional American school settings. It may he that such settings are theevolutionary outcome of a social system selecting for effectiveness with childrenof the dominant culture. The goal of new research should he to examinesystematically the degree of match or mismatch between characteristics of.4merican families and the characteristics of schools. Second, research shouldemphasize, even more than it has in the recent past. patterns of variables in the

Social and Developmental Context 23

families of readers (and, of course, the families of nonreaders). The discouraginghistory of attempts to establish aptitude-treatment interactions suggests that weneed, on one side, more intensive case and diary studies of reading and, on theother, thoughtful experiments in the natural settings of homes in order toilluminate the genesis and early history of the preparation for reading.

Cross-national and historical studies of early reading instruction, as well,may contribute to cu: understanding of what environments facilitate successfulreading, but the most productive studies will probably be directed toward threeinterrelated questions:

(1) What variations exist among American families in their preparationof children to read? (For example, are there systematic variations in theway children are led to focus attention on a task or a game'?)

(2) What changes can be made in the lives of families that will makereading easier and more rewarding? (For example, will simply makingmore hooks available to families change the preparation of children?)

(3) What changes can be made in early instruction in reading that willtake optimal account of the diversity of American child-rearing prac-tices? (For examples. will programs that encourage more active ap-proaches to reading improve the relatively poorer performance of boys?Is television being adequately exploited?)

A specific exaoiple of an appropriate experiment would test the hypothesis

that flooding the home environment will increase reading skills. It can beassumed that the correlations between charac, eristics of the home and readingachievement represent a two-way relation: it would he important to learn how

much effect we might have by modifying the home environment. It would also

he useful to know if we could reduce or remove the relation between readingachievement and the aggregate SES characteristics such as income, occupational

level and parents' education by flooding homes with materials or by otherstrategies for encouraging reading.

b. The use of the child's experience. Beginning school represents, for

many children, a significant discontinuity. They are suddenly placed in a settingthat operates under a new set of rules and asked to perform tasks that many ofthem find strange and unconnected with the world beyond the classroom.

Strung support should he given to research that trill explore new ways of relatingthe child's out-oischool experience to Ids first tries at reading, particularly forchildren whose experience does not fit the traditional expectations of theschool.

There is some anecdotal evidence (especially from the new English primaryschools) of the effectiveness of programs based on the child's "own voice" and,


more, an accumulation of strategies. materials, and ideas about what a classroomtied to the child's experience should look like. We were impressed by the widesupport given such a program by observers as different as the radical schoolreformers and the bicultural educators. In order to construct sensible changes inthe environment of early learning, at least two lines of research are indicated:

(1) What is the world of American children like in the years beforethey enter school? We need studies of family structure, communicationpatterns, and allocation of interest in order to arrange effective patternsof connection to school. Again, the variety of American culture re-quires that such studies represent the variety of family cultures.

(2) What are effective ways to induct children into school? It is notnecessary for all American six-year-olds to enter first grade on theWednesday after Labor Day. Nor is it necessary that the transition fromhome to school he abrupt. Research is required to understand theeffects of staggered admission to school, of parental involvement incurriculum design and teaching, and the use of schools as educationalcenters for their communities.c. The measurement of competence. Elsewhere in this report we discuss

the problems of assessing the abilities of children. Measures of general concep-tual ability perhaps based on Piaget's observations measth.- of attention,and measures of the comprehension of oral language can apptopriately anduselidly he devised JOr the child who is not vet in school.

d. The media and reading. Surely the most important development inAmerican education over the last two decades had been the introduction ofuniversal television. We know remarkably little about the effect of the change ingeneral, and with a modest exception (preliminary studies of the impact ofSesame Street and The Electric Company). we know nothing about the effect oftelevision on reading. Systematic studies of the effect of the media on reading,and research on new modes of television teaching must take a high place on theInstitute's list :)1 priorities. The possibility of interactive television where aviewei can communicate back to the source by means of cable transmissionshould not he overlooked. The active involvement of parents and peers mightalso increase the effectiveness of televised instruction.

e. the adult learning to read. An unusual opportunity exists to study thepreparation of readers and the interaction of motivation, setting. and compe-tence. Open-enrollment colleges have discovered that some of their students donot read fluently and that some of them can scarcely read at all. Retrospectiveresearch on the young adults who first leant to read fluently in college mar tellus more about the social context of reading. ('artful studies t0.* how suchstudents learn to read as adults will help us to understand teaching and learningstrategies and open as well another window on the definition of practicalliteracy.


f. A note on method. Research modes vary widely in studies of education.Without diminishing the importance of laboratory studies, we recommend thatmore use be made of intensive case studies (for example, of the families of verygood and very poor readers), anthropological type observations (for example, ofdiffering ways of routing children from home-culture to school culture), andlarge-scale field experimental studies, particularly those that involve differential:rapport to families of preschool children.

2. Characteristics of Teachers and ClassroomsDespite the increasing weight of evidence from large-scale survey research

that school factors make relatively little difference in determining achievement,we do not believe that the search for potent influences in the school environ-ment should he given up. On the contrary, we believe that hypotheses now exist,in the empirical results of small-scale studies and in theoretical models of schoollearning, that should be tested more systematically.

There are several reasons for our insistence that such research is importantand worthwhile. First, although home influences :ein now, and may remain, thesingle largest correlate of school achievement, they explain only a small part ofthe total variance in reading skill. Se':ond, the standardized norm-referencedtests used in these surveys of achievement are designed to differentiate amongindividuals rather than among environments. Tests should be developed to assessmastery of particular skills defined either on theoretical or functional criteria.We expect that such measures will aid the search for important school, curricu-lum. and teacher influences. Finally, large surveys of school achievement havegenerally used questionnaires to assess school and teacher characteristics. Mostquestionnaires produce superficial and sometimes even invalid data about thecharacteristics of schools, teachers, and classroom behavior, thereby making ithard to detect differences. Instruments using direct observation of behavior andthat are based on hypotheses about significant variables and configurations ofvariables are needed.

In Addition to refinement of large-scale studies, the following aspects ofinstruction, classrooms, and schools merit further investigation.

a. The stability of teacher effects. One assumption underlying research onteacher characteristics and, olrimntely. much school-based research. is that someteachers are ronsistently more effective than others. This assumption should heexamined. -Iwo approaches are reasonable. First. it may he easier to detectsignificant characteristics of instructional settings in the primary grades becauseschool influence is likel to he more potent with younger children or becauset ew schools have consistent programs throughout their classrooms. Second, weshould look at the consistency of classroom "gains- one year to the nextto obtain estimates of the "maximum amount of effect that might be due todiff_rences among teachers. One tentative :et of findings using this approachsuggest: that the largest classroom difference: were in the first grade, the next

20 tinguistic (')nuuiatiti. I'rspectires .1i,r Research

largest in second grade, an.. s on in decreasing magnitudes until no differencesremain at the fifth grade.

h. The place of structure in classrooms. (el tam instructional modelse.g.. the structured, acadernic...iyoriemed curricula appear to he significantly-more successful in raising very earls reading achievement than others. Thecommon elements underlying this success should he sought. and the approachtr:...;1 should he studied longitudinally to determine it the ;ii.coss is retainedduring the later grades.

C. Gains during the school ,rear and during the summer. It has beensuggested that mitionty children tend hi regress during summer vacations.whereas middle-class children may actually gam. It this observation could heestablished as valid for rending and linguistk competence, it would have obviousimplications tot potential intervention programs.

d. the effect of peer tutoring. Research on the roles of peers. particularlyas tutors. should he continued and expanded. .1 lie tutorint7. process should alsohe defined to include observations of the value of sharing. not merel acquiring,new knowledge and skills: there may he too much competition too soon in thet radit tonal elassror.

e. The characteristic of open classrooms. A list of variables has been citedas important m [Molina! or open educational settings in }loth 1.ngland and the1 lilted States: quiet ,paces. emphasis on writing as well as reading, ignoringimmature spelling. high relevancy of materials. :loss -age grouping. these oommot' , represent the same qualit of data as case studies can yield, andwai lam more systematic resealch.

I. Opportunity and moliralion. One model of school leaning that offersfly potheses for Rather testing has heen developed 19(0). It is

possible that "PPol utility and motivation. two of carroll's variables, may under-lie and extllatn the etleotventNs of a variet of successtul programs. Some of thedifferences found III pievious research on progiams of teaching reading may beasciibable lo different amounts of tine spent by the child on some aspect of thereading process. And sk:hool-pioduck.sil differences m explain theet; ectiveness of plocedures designed to mk.rease attention. if motivation count))sthe intensity of dIllds ;1110111i:11 at an given moment and the ahilit tosustain attention ovei long pehods ut time One suggested approach for testingthese Iry potheses Is to hold curriculum constant and to examine individualvariation inn the actual time spent attending to reading and learning to read.I. 'sing this appioach. it should he possible to estimate the effect of the variableson child out,:omes uncontaminated by cUrrieulum Lid terences.

wo dltleretrt models ale sul.2ested for Sudi research. One ,..4)nsKtssullies of successtul rearleiN successful teachers and successful schools. Anothermodel entails comp:it alive studies of plograms where one or more variables issystematically varied. In the latter case. serious Held development is needed first,followed bv, field experimentation.

Social and 1)0ielOtitliVnt.11 111111.`1,1

Finally, Iv(' sugwest that should not initiate .titrtlier large-scale. cross-Wel ir reScilre 11 deigned 1r) 86111111e' e(111.(11 relt11101IS It home,and peer characteristics on the one hand and ycholastic achievement on theother. closN-se,:tional itio,ei are applopoate for assessment aLtivities, nit forisolating causal mechanisms.

Characteristics of the Reader

Research relating 1,, pupil chaiacteristics has moved in many differentdnections and some ploblems have been much moie lulls explored than others.Among the most piominent issues the ;ilea of pupil characteriic.: and readingale the following.

1 (haracteristics of children who read early. Longitudinal studies of earlyreading 14c ahead been undertaken on a small scale and have raised soloeinteresting kpiestions. Some of these studies suggest that early reading 1% mostbeneficial till children of average intt:Ilectual ability and least beneficial tochildren of beller than avelage intelligence. These preliminary findings need to

he expii)ted 111 nin1/4:11 mone detail because of their potential relevance to thewhole issue of early reading instruction.

h. (liddren's concotions of reading. It it is tune las we ht.slieve it is) thatthe 1/4.11ild's river,uiI ,attitude towards reading plays a part in his performance, thenwo should know how and what he thinks about the process of learning to readand !bruit the nature of letter's and words. Some preliminary ro,:ults in thisdomain suggest that young children believe that one can only read at a certainage, and that Noting children have a very vague conception of what letters andwords ate.

c. Cross-cultural case studies. Descriptions of pieieading experiences ofchildren in different 1/4.ountries, together with information about their formalreading instruction and reading achievement. could he of interest. Obviously,vast cultural and language differences would impose limitations upon these

alley might providecompaosons, hut the could nonetheless he suggestive.insights into the cultural context of It...minig to read that would not be possibleto obtain 111 an other wa . Moreover, trout a practical pinnt of view, theremight well he teaching practices and materials used in these countries that are ofvalue and which might he adapted to the teaching of American children.

d. Social p.svcholo of the classroom. Much inure work is needed oil theclassroom as an important social unit. Some studies suggest that particularcombinations tit teacher ab..1 pupil personality characteristics are most effica-cious in learning, It is probably also true that certain combinations of childrenwalk better than others as a learning group. Within any school system, particularclasses are kinovii either -good- in -bad- it seems to he, in part at least,independent of the teacher who has the group. Research in this domain mightmake it possible one day to select children when they enter school in such a way

Linguistic Communication: Perspectives pp Research

as to fit them into a group that would maximize their learning propensities and

minimize factional controversies.

e. ,1 note on method. In studies dealing with pupil characteristics, a rangeof methods is appropriate and should he applied. Observational studies ofparticular children and of classrooms. ethological studies of children on theplayground or at home should have their place along with the inure carefullycontrolled laboratory or classroom experiment. It is important to provideinvestigators wide latitude in getting at the problems they are attacking. Differ-ent levels of rigor can he expected in a field where there are many differentproblems at many different levels of conceptualization and investigation.

1. Diagnosis understandilf the child's competence. It is important thatclassroom teachers, as well as clinic:1 specialists, he able to make sensitive andusable diagnosis of a child's preparation for reading and his growing competence.Some of the devices and instruments needed for understanding what the childknows already exist others will awed considerable research before they can bedeveloper. Both circumstances should he flagged.

However. and the point cannot he too strongly made, the overridingproblem in individual diagnosis is not in instrument development. It is rather inthe attitude of teachers and other adults toward individual differences amongchildren. It' we could somehow convey the notion that diagnosis and teaching areinseparable, we might reduce the need for large-scale efforts in instrumentdevelopment and rely more on the intuition and sensitivity of experiencedteachers to evaluate the preparation. competence. and needs of their students.

4, Influence on Reading of Dialectal VariationCurrent sociolinguistic data suggest tl it many children speak a well-ordered.

highly structured dialect different from standard English. This fact may haveimplications for both test cont ruction and the development of reading skills.

The importance of considering different language systems in test construc-tion can he demonstrated in four areas of potential difficulty for the speaker ofa non-standard dialect of English: in the content of the test questions andexpected responses. in the verbal style required by the test. in the non-linguisticfactors inherent in the testing situation, and in the linguistic aspects of the test.Substantive biases in tests can include specific vocabulary items. culturespecificpictures used in vocabulary tests, as well as culture-specific information

questions.

In addition to receiving vocabulary and information penalizations. a childmu in the mainstream can he hampered in his test performance by the verbalstyle required by a test. For example. in some subcultures, a standard ofarticulate description might he specificity and brevity rather than the testmaker's usual standard of generality and quantity of description.


Situational factors can also work against speakers of non-standard dialects.The activity of being tested can he intimidating to a child and hamper hisperformance; the realization that he is expected to produce answers according tonorms not characteristic of his own culture may cause him to resist the testing oreven to refuse to participate. Moreover, forced interaction with an adult whospeaks another dialect and is of a different race can have adverse effects on thetest performance of children.

Purely linguistic aspects of the tests may also cause problems. We do notknow to what degree speakers of other dialects, particularly young ones, com-piehend standard English; different semantic connotations and denotations ofwords, and perhaps different implications and presuppositions of sentences indifferent dialects, may interfere with comprehension.

Given this state of affairs, what solutions can be offered? A good first step isto consider alternatives to deficit theory when explaining cultural group differ-ences in test performance. When differences appear, researchers need to explorethe different conditions which do, or do not, produce the variation. Forexample, the following line of questioning might be pursued: Why does the samequestion asked by an experimenter of different kinds of people receive differentkinds of answers'? What do respondents perceive the questions as really asking?What, in the respondents' culture or linguistic background, has made theminterpret a question in a specific way? When the experimenter does not receiveresponses he considers appropriate, does he then ask himself what conditions areabsent which serve as cues to elicit the appropriate response?

All of these questions are relevant when we consider black children. Poorreading performance by black children continues to plague the classroom. Manyexplanations have been advanced for this state of affairs; a recent one revolvesaround their language. There are two views of black children's language asdeficient and as different. The deficiency view is that the language of blackchildren is an inferior form of standard English. The difference view is that blackchildren's language is an expression of their culture and a viable system oflinguistic communication, equal in dignity to any other. Proponents of thedeficit theory would eliminate the black English vernacular and replace it withstandard English. Thus the source of the problem would he removed. Thosefavoring a different theory would change the schools and methods and materialsfor teaching reading in order to accommodate to black children's language.

With regard to interference with reading when one is a non-standard Englishspeaker, two kinds may exist: phonological and grammatical. The data onreading interference attributable to phonological characteristics of the blackEnglish vernacular arc by no means clear. Most of the research suffers from somemethodological flaw, so that the question of phonological interference in readingmust remain open. The different syntax of the black English vernacular couldinterfere with reading ( csrcially with comprehension) in two ways: interference

30 Linguistic Communication: Perspectires jOr Research

could arise where a standard English sentence is interpreted as a non-equivalentsentence in the vernacular, or when a speaker of the vernacular has to translatehis own speech into standard English. However. the plausible claim that this"translation process" interferes with learning to read has not been firmlyestablished.

The matter of dialect interference with reading is a complicated one,complicated primarily by the fact that children particularly older childrenare capable of code-switching; thus, they cannot he said to be monodialectalspeakers.

What can he done? Clearly, there are black vernaculars that differ fromstandard English. The extent to which they interfere with reading acquisition isby no means clear; it is an area requiring further research. The lack of solidevidence here militates against a positive recommendation that black children hetaught reading in their dialect. Further, some black parents are opposed to sucha procedure. although good evidence of advantages might change their minds. Anappropriate piece of research here would involve teaching children to read withdialectal supports perhaps with dialect readers and then comparing theirreading with children taught to read by standard English readers. Such researchshould study children at different ages and stages in learning.

Speakers of a non-standard vernacular may have difficulties with currentinstructional offerings in reading, and writing. but speakers of a different lan-guage can he completely frustrated. The problem is that instructional materialsand procedures with special attention to ethnic minorities comparable to thereaders and trade hooks available to children who speak standard English are

nut available. In this situation, the natural objective to set up is the productionof an array of instructional offerings which are responsive to the various learningcontexts within which minority children grow up. We suggest this is an area forfurther study by NIE.

B. THE PROCEVSES MNG AND WR/T/NGWe do not know the degree to which an understanding of the fundamental

cognitive and perceptual processes in reading can contribute to solutions of thisnation's literacy problems: some experts hold that reading difficulties havepiedommantly social and political origins. But it seems likely that detailedknowledge of the abilities that are needed in order to read effectively will makeit easier to diagnose and remedy the specific reading difficulties that occur inany given case, regardless of their social origins.

I. Basic Literacy

Basic 'eiacv is a complex and purposeful activity which draws on manypsychological processes and requires skills in a number of component abilities.

Processes of Reading and Writing 31

inadequate development 0- use of any one or more of a number of componentskills may impair the ability to retrieve the spoken word from the printed page.The processes that underlie these skills, tiPrefore, all constitute potential sourcesof reading problems. We group the component skills and processes under fourheadings pattern - vision skills, visual-phonic skills, visuomotor skills, and cogni-Live skills and list what appear to he important areas of inquiry and applica-tion in each case,

a. Skills of pattern vision. A child must learn to distinguish and identifyindividual letters and letter groups over a wide variety of type styles and scripts.lie must be able to recognize the spatial order in which the letters stand, and totranslate that spatial order into a temporal sequence, it' he is to be able to soundout syllables, words and sentences. And he must learn to recognize the patternsthat are formed by common letter-groups, without dissecting them into theirconstituent letter-shapes, if he is to progress beyond the most primative stages ofreading,

Although there have been many studies of the legibility of different stylesof print, results have provided little knowledge in the areas where we need itmost: we need more knowledge about the distinctive features by which childrenlearn to hlentift syllables and mmls: about how they learn those features: andabout how the process of learning may be facilitated. Research methods exist forthe study of these questions, sonic research has been performed, and more isneeded.

As skills develop beyond the most primitive levels, the reader's translationof text into speech is informed as much or more by the context and content ofwhat he is reading, as it is by the individual patterns at which he is looking, andwe need to know inure about how this process works and where it fails. Severaltheoretical models have been sketched to de,...iribe this process, but none hasbeen developed to the point where it can either be rigorously tested or fruitfullyapplied to diagnostic and remedial programs. Specific efforts in these directionsshould he encomaged.

Visual-phonic skills. Writing is a system for encoding ine;.ning graphi-cally. Reading is the translation from writing into that form of hInguage fromwhich the reader is already obtaining meaning. Phonemically oriented writingsystems. including alphabetic systems like English, permit speech to he frozen invisible marks and subsequently defrosted back into speech. One basic problem inwading phonemicized writing might be conceived as discovering and using theinks of the graphic code to recover the original spoken language. This concep-tion of reading pre-supposes that once the spoken-languag form is recreated, itsmoaning will he evident: the reader who has mastered this skill will be limitedonly by his ability to comprehend the spoken language.


To view written English aq merely a system of letter-sound correspondences,however, tends to obscure a significant property of alphabetic writing. Alpha-betic writing systems encode spoken language not only at the phonological level,but also at the lexical, syntactic, and semantic levels. A good reader seldomresorts to phonemic decoding, although he can do so when he encountersunfamiliar words. In decoding printed English, a reader can analyze at the levelof individual letters, letter clusters (digraphs and diphthongs), spelling patterns,syllables, morphemes, phrases, clauses, paragraphs. No single level of codeanalysis characterizes the task of extracting meaning.

The process that is often called "decoding" written English seems to havethe following features:

(1) The end purpose of the process is the extraction of meaning, notthe mere pronunciation of speech sounds. (Even at the phonologicallevel of decoding it oan he argued that the reader is uncovering thephonological "meaning" of the code, although our main concern here iswith the semantic meaning ot' the code.)

(2) The mental activity of the reader in decoding written English, atwhatever level of analysis, is an act of problem- solving requiring somedegree of reasoning skill.(3) Context is as important in decoding written English as it is incomprehending spoken English. Contextual clues of all kinds phono-logical, lexical, syntactic assist the reader at all levels of code analysis.

Before a child is ready for this learning experience, he must have certaincomponent skills that the experience can build on. In addition to patterndiscrimination and control of eye movements, there are also auditory skills thatmust he acquired prior to learning letter-sound correspondences. An ability toanalyze speech into consonants and vowels may be particularly critical forchildren who are learning to read an alphabetic form of writing. If a child isunable, for whatever reason, to hear constituent phonemic elements, he will bemystified about what the printed letters arc supposed to correspond to. It hasbeen established that this ability is important; what to do for children who lackit is still a matter for debate. One approach postpones the problem by usingwhole syllables (or even rebuses) to introduce the idea of reading at the level of asyllabary, rather than an alphabet. Another approach develops special pro-grams e.g., word games based on such auditory contrasts as man-can-tan-fan,man-mad-map-mat, and so on to train a child's ear to hear these segments ofthe speech flow. Studies directed toward the resolution of this argument shouldcontinue.

It is commonly assumed that familiarity with the names of the letters is anecessary preliminary to the teaching of letter-sound correspondences. I lowever,children have been taught to read before they learned the alphabet. It is

convenient for a reading teacher to he able to refer directly to particular letters


by name, but most letter names differ from their phonological realization inspeech, and the difference may confuse some children,

It is also commonly assumed that a child will he familiar with the variousEnglish words he is expected to identify from their written forms, Childrenwhose first language is not English will not have mastered the "familiar words"they are expected to identify, and may even have difficulty hearing the compo-nent sounds in English speech, The special problems of bilingual and biaialeetalchildren are discussed elsewhere we consider here only children who alreadyknow and can pronounce the words whose alphabetic representations they areexpected to learn,

Three stages of learning letter-sound correspondences are generally distin-guished, The child must I) master letter-sound correspondences, 2) learn to"blend" these sounds into syllabic units, and 3) learn to deal with polysyllabicunits,

The letter-sound correspondences that occur in English orthography arevery complex. Twenty-six letters of the alphabet must he used to representbetween forty and fifty phonemt s (the number depending on the precision ofthe phonological analysis). The mapping is many-many: in different contexts aletter may have different phonemic realizations and a phoneme may be repre-sented by different letters. Attempts to estimate how many letter-sound corre-spondences there are have produced numbers over two hundred, of which nomore than 2S percent can reasonably he considered special cases, or "excep-tions," How many of these correspondences a child should learn can only beestimated from studies (not yet undertaken) of how much use they are to him.

If one imagines a child who knows all the rules relating letters to sounds, hewould still not he a reader. He would still have to master the skill of combiningthese speech sounds reading teachers call it blending before he couldconvert the string of sounds into a recognizable word. Compared with learithigthe rules of correspondence, blending seems a logically simple task, yet teachersfind it is one of the commonest sources of frustration for a beginning reader. Achild may he able to look at dog and say Al& "dee, oh, gee," or "duh, aw,

lie may even to able to say these sounds in rapid succession, yet hearnothing in the resulting trisyllable "duhawguh" that reminds him of the mono-syllabic dog. Blending is the converse of the auditory analysis of a spokensyllable into phonemes: some workers have seen a close relation betweendifficulties of synthesis and analysis. The synthetic process may be more com-plex, however, because the child must not only blend speech pounds into longerunits. but must also take into account various silent letters (e.g., the final e inhope) or letter combinations that mark which particular letter-sound corre-spondence is appropriate (e.g. whether o is to he sounded as ah or oh ).

The third set of skills that a beginning reader needs relate to the formationof polysyllabic units. Phonic rules will enable a child to sound out syllables, but


if his analysis of a polysyllabic word leads him to the wrong syllabification, hewill produce an incorrect pronunciation. For example, nation may he seen as atrisyliable,"nah-tie-on," rather than a disyllable, "nay-shun." Problems posed bywords built of two or more morphemes, such as "running," differ from thoseposed by polysyllabic morphemes like "nation." It is usually assumed that achild who reaches this problem will be able to solve it for himself; relatively littleanalysis has been devoted to problems at this level of difficulty.

Such analysis of the decoding operation lays bare its considerable complex-ity; the wonder is not that some children fail to master it, but that most childrensucceed. Not only must this complex system of rules be internalized; it mustoperate automatically at rates equal to or gr6ater than the rate's of normal speechif the child is to devote attention to the meaningful content that gives point andpurpose to reading.

How should such skills be taught? A variety of answers have been offered,and repeated attempts have been made to evaluate the effectiveness of com-peting proposals. Consider two cur trusting philosophies. A teacher who decidesto teach letter-sound explicitly could demand the memorization of two hundredletter-sound correspondences, of complex rules needed to determine whichcorrespondence is involved in any particular instance and of rules for seg-menting long words into the syllables to which the system of rules can apply.Alternatively, a teacher who decides that children should associate spoken wordsdirectly with printed .etter-patterns (in much the way Chinese children learn toassociate spoken words with written characters) would abandon the alphabeticprinciple of word identificatioi on the basis of phonemic analysis. In suchextreme forms, of course, neither philosophy is attractive. A child who recitesrules to himself before he pronounces a word will he slow in gaining the speedrequired for competent reading. A child who learns every word as a new taskmay start more rapidly, but will he helpless when confronted by novel letter-patterns. In practice, therefore, instructional methods normally include bothapproaches. A child may he introduced to the idea of reading by learning a fewwords of relatively simple and regular spelling by the "look-say" method. thenhe given a few of the phonic rules chosen to be the most productive in increasinghis reading vocabulary, but not so many as to slow his reading of simple texts"basal readers" designed to match his growing competence. Curriculum designcenters around the relative emphasis to assign to the two strategies; at presentthe fashion is to be a little closer to the phonic-rule approach than to thewhole-word approach. The question for which we still need ei definitive answer ishow far the teaching of these initial skills must be carried before the child beginsto generalize fir himself:

Another set of issues intrudes when we consider learning to write. Readingand writing are obviously mutually supporting skills, but teaching strategiesrange all the way from treating them as independent skills to the claim that


writing should he taught before (or instead of) reading. If spelling niceties areinitially neglected, and if problems of motor coordination are bypassed byproviding block letters or a typewriter, writing can he taught at an early age. Atpresent, neither learning process is well enough understood to enable us toevaluate claims about their interrelations or about instructional tactics based onsuch interrelations.

The teaching methods and materials that are presently available seemadequate to enable most children to acquire at least minimal skills in decodingwritten texts. Even children who for any of a variety of reasons, have difficultywith reading can be taught it' the teacher has sufficient time and is sufficientlypatient to take it the fact that they are often neglected is more a consequenceof the classroom situation than of any ultimate incapacity on the part of somechildren.

Further large efjOrt to develop new versions of the reading curriculum alongthe familiar lines of curricula currently available is not advisable at this time,although pilot studies of alternative approaches should always be possible.Higher priority should be assigned to obtaining better information about whatregularities of English orthography children are actually teaming under theexisting systems, and to the development of curricula better tailored. to theprereading experience and abilities of children from minority groups.

The length of this discussion of visual-phonic skills should not be inter-preted to mean that they deserve special emphasis from research workers. On thecontrary, it means that the analysis of these skills has already received moreattention, so that more is known about them, than have other aspects of thereading process.

c. Visuomotor skills. The fine distinctions that are needed to recognizeletters and words can he made only in a small central region of the eye, calledthe Prea. The fovea is wide enough to encompass only a few letters at a time, sothat the eye must move from point to point in order to read more than theshortest text. (f low the rest of the eye, and peripheral vision, contributes to theprocess is nut now known). The eye movements that are most important inreading are jumps that hrii some new point, initially seen only in peripheralvision, into the tOvea. About 4 such jumps, or saccades, can he made per second.A beginning reader must be able to direct his gaze successively from one point toanother point close by as he spells out the text by letters or small groups ofletters. As his pattern-recognition skills improve and as his expectations aboutwhat the text will say improve), the points to which the eye moves can befurther apart; he has to sample the text relatively infrequently. But in eithercase, controlled and purposeful eye movements are at the heart of skilledreading. In addition, the reader must he able to progress along the line withoutlosing his place, horizontally or vertically. And finally, he must be able to fit thesuccessive glimpses by which he has dissected the text hack into a single ordered

36 Linguistic' Communication: Perspectives for Research

structure, even though he frequently returns to look at material that his eyeshave already traversed.

The eye movements that are made in the course of reading are therefore tobe viewed as necessary skilled behaviors by which information is gathered; theyare also to be viewed as sensitive indicators of what expectations the reader hasabout the text, and what he needs to know. Training procedures that attempt toimprove the efficiency of the reader's eye movements deserve study, but thesuccess of such efforts must be limited by how much knowledge the readerbrings to the task, knowledge which permits him to fill in between widespreadglances. That is to Say, most of the attempts to give poor readers practice inperforming eye movements per se have shown little promise: training the readerin efficient search behaviors is of little use unless he already has a sufficient baseof knowledge about the world, and about the redundancies of language, to 'akesuch search behaviors profitable. These search-training methods will have verycircumscribed use until they are brought into the context of well-developed andexplicit theoretical understanding of the reading processes they serve.

Although there is a long .history .Its studies of how eye movements con-tribute to the reading process, we need detailed tests of theoretical models ofhow the reading process proceeds. At present, the measurement of eye move-ments is cumbersome, expensive, time-consuming, and interferes with thenatural set of the reader. To some extent, we can study the contribution of eyemovements by "simulating" them (i.e., by presenting motion pictures of textwhich present sections of each line successively), and by other indirect methods,but these methods cannot be wholly satisfactory. We need to develop better andless obtrusive methods of recording eye movements, and more rai:(1 and eco-nomical ways of analyzing them. We need such methods to assess readers'performance; to evaluate the effectiveness of teaching materials (including pic-ture hooks, television, and animated displays); and to use in training proceduresthat make the display itself contingent on the eye movements that the readerexecutes. In any case, the recent development of recording devices that aredirectly coupled to computer-generated displays, and the simulation of eyemovements by the use of motion pictures to provide successive views of texts orscenes. have resulted in much new information about the basic nature of thereading process.

Many of the tasks involved in acquiring basic literacy are difficult andunpleasant: mak:, a succession lat. tightly controlled eye movements, which thebeginner must do at least in the "sounding out" stages, is an unnatural andprobably aversive way of using one's eyes; the ability to "combine" or "blend"phonic elements into syllables and words does not come easily (and in somecases takes an inordinate amount of practice to acquire at all). We might he ableto make at least the initial stages of learning to read a great deal easier (e.g., byusing letters and words that are easy to distinguish and to remember; by using


language, concepts and stories that are already familiar to the reader; by usingpictures and text in an optimal balance to give the reader an advance idea ofwhat the text is about without destroying his interest). More research is neededon ways to make text as easy for the beginner to read as possible, anddevelopment and dissemination procedures (including television and comics)should be provided to bring the text to as wide an audience as possible.

By minimizing early frustrations and difficulties, and by getting the child tothe rewards of reading as rapidly as possible, it may be feasible to bypass someof the present sources of difficulty; it may be possible to reduce other difficul-ties to the status of less salient incidents; and it may be possible to defer theleast immediate problems, and to provide intensive instruction in those lettersand words whose teaching cannot be deferred and which experience has shownto be particularly difficult.

We cannot at this time be certain about the long-range effect of deferringdifficult aspects of reading. An analysis should be made of research resultsproduced by the 1TA experience, and of other field experiments related to thisissue, to determine the extent to which initial gains are sustained after the readergraduates from the training alphabet to standard textual material. If such gainsdissipate entirely when the student finally has to face difficulties that have beenbypassed during his early training, the expense of the effort to defer thosedifficulties must be justified on grounds other than the requirements of aneffective literacy program.

In his briefing to the Study Group, Coleman asked for gathering all word-

use data into one source, and spoke about his "engineering" tables based on suchdata. We should determine what the actual predictive efficiency of such tables is

(and going somewhat beyond his proposal try to induce what the distinctivefeatures are and exaggerate them, where possible). One would also like to seesome attempt to scale stories (or themes or plots) in terms of their interest atdifferent ages. Note that these studies need not be done with written texts; they

could be done with animated cartoons or movies, and presented to preliterate

and illiterate populations.

d. cognitive skills. Over the years there has been a continuing controversyover the cognitive abilities involved in learning to read. In the past some haveargued that a mental age of 6 5 is necessary before a child can profit fromreading instruction. In more recent years. workers in the Piagetian tradition haveargued that children should have acquired concrete operations (e.g., an ability toreason syllogistically) before reading instruction is undertaken.

The rationale for this position is that reading, from the very outset, is acomplex cognitive process that enables a child to acquire meaning from thewritten word. From this point of view, even the acquisition of letter-soundcorrespondences requires an understanding that letters are arbitrary symbols. In

3$ tinguivti Communication: Perspectires for Research

the absence of such understanding,. it has been argued, the child may acquirehabits that are detrimental to fluent reading and that might have to be unlearnedlater.

In support of this position are some studies showing that children who aretaught reading early have no lasting advantage over those who are taught at alater age. Workers who support this "readiness" position also point to countrieswhere reading is usually not introduced until the children are seven years old. Insuch countries (e.g.. Russia, China, and certain European countries), readingdifficulties of the kind encountered in America are said to he relatively rare. (Itshould he said. of course, that the writing systems of these countries mighteliminate many of the cognitive problems posed by English orthography.) In anycase, the position of the cognitive readiness group is that formal instruction inreading should he introduced no earlier than the first grade.

On the other side of the controversy are those who believe that readinginstruction can, and perhaps should, he introduced earl,. in life. The argument isthat basic literacy, the connection of printed letters with spoken sounds.requires no more in the way of cognitive skills than discrimination and associa-tion. Since even very young children possess these abilities, there is no reason fornot teaching children to read at an early age. either at home or in a nurseryschool or at a day-care center.

The real problem of teaching reading. according to the discrimination-association position. is to make the task sufficiently simple for young children.A number of large-scale research projects have been under way for quite sometime with iust this aim in mind. Their goal is to calibrate the difficulty of letterdiscriminations, of letter - sound correspondences. of blends, and so on, so that allthe various analytic reading skills can be ranked with respect to difficulty fromthe child's point of view. Once this calibration is complete it should he possibleto design reading programs that would move children from the simplest to themost difficult basic literacy skills in a systematic way.

In support of this "no readiness" position on reading instruction, examplesare often cited of individual children who have been taught to read early.Likewise, studies in which children have been taught to read or attain basicliteracy skills early by way of particular programs of instruction are also referredto as evidence that reading can be taught early. Experiments with Distar (SRA)and with ITA are of this sort. Other evidence comes from children who attendschools where reading is introduced before first grade. such as the Montessorischools. These schools provide additional evidence that reading can be taughtbefore a specific age.

As in every controversy, differences in substance are often exaggerated bydifferences in definitions. Those who argue for the delay of formal instructionare at the same time quite vocal in advocating a variety of prereading experiencesand activities. They propose that children experience a language-rich environ-


ment, that they be read to, that they become acquainted with words and letters,and that they have some experience in printing and writing. What the "readi-ness" group Nall) opposes is putting young children into formal teachingsituations where they must learn formal rules of phonics, spelling, and so on.

In practice, the "no readiness'' group is really not far removed from theposition of the "readiness" advocates. Programs for teaching reading to youngchildren inevitably adapted to their characteristics as learners. Letter-soundcorrespondences can he taught in a multitude of ways, many of which areinteresting and enjoyable for young children. In good programs of this kind, theneeds, abilities, and interests of the children are used as a starting point forintroducing basic literacy skills.

In view of the above discussion, some conclusions and suggestions forresearch can he made. First, some clarity has to be brought into the discussion ofreading instruction. Much of the controversy between the readiness and the noreadiness groups derives from the definition of "formal reading instruction." Itseems incumbent upon any research worker in this field to be explicit about thekinds of instruction he is undertaking and the extent to which it is modified inactual practice.

It would seem appropriate Pr the :VIE to support research into the cogni-tive processes (reasoning and perceptual abilities) that are involved in basicliteracy skills. In addition, it would seem appropriate to support studies dealingwith the analytic component of basic literacy skills, and which aim at calibratingand ranking these skills and building them into a comprehensive early readingprogram. Researchers working on both of these problems should be broughttogether to share their findings and views, and perhaps to arrive at commonlybared convictions and joint research ventures.

Comprehension of LonuageThole i, Rio intrinsic difficulty (although there may be many practical and

technical problems) in testing or measuring whether people can comprehend oralor written messages. Comprehension is tested by determining whether people,after receiving a message, can answer questions about its contents or implica-tions. or can carry out its instructions. But there is a long leap from measuringcomprehension of a message to discovering the thought processes a person mustgo through in order to achieve that comprehension.

Standard forms of school instruction in reading and language arts do muchto facilitate the growth of language skills in elementary and secondary schoolchildren as they mature. These instructional methods are based on a wealth ofpractical experience in teaching, but they are probably most effective inhandling average, normal children, because it is such children to whom most ofout experience relates. Standard instructional programs are certainly not basedon any extensive analysis of the processes underlying the skills, because suchanalysis still remains largely to be carried out.

40 Linguistic Communication: Perspectives fir Research

With a better theoretical grasp of the reading comprehension process tosupplement conventional wisdom, we could devise more powerful diagnostictools to identify with accuracy the weak and strong points of individualstudents the processes in which they are competent, and those in which theyneed additional help and training. With better diagnosis, our deeper understand-ing of process can also lead us to more effective teaching and learning techniquesthan those we now possess.

During the past decade there has been a great burst of new research activityin psychology and related disciplines aimed at understanding complex humanthought processes, including processes for comprehending language and problem-solving processes. Oi reason for this burst of activity, and the consequentprogress that has been made, is the availability of new tools of researchcameras for detecting and recording eye movements, and digital computers, Tomention just two examples. These tools have provided us with new ways toidentify successive steps in human thought processes, and to formulate modelsof those processes that will permit their operations and pathologies to beexplored and tested. Studies of successive foci of attention in reading tasks,using eyemovement cameras and thinking-aloud protocols, should continue toexpand on our knowledge of these processes.

How language conveys information from one person to another is no longeran unpenetrated mystery. Enough is known abt ut how language carries mean-ings to permit us to write programs for computers that accept instructions innatural language, and other programs that solve the problems posed by thoseinstructions and carry out the tasks they set. These programs "understand"language by the same criteria we use to assess human understanding they cananswer questions on the basis of what they have read, they can carry outinstructions, they can solve simple problems. Moreover, these programs are opento analysis, so we can study exactly what processes enable them to do thesethings, and compare their processes with the behavior of humans to whom thesame tasks are presented.

Computer simulation of human cognitive processes is one of several promis-ing techniques for modeling the methods used by people to perform complextasks. It is well adapted to diagnosis of individual differences in performance andto tracing the differences in process that underlie them. For example, in learningalgebra, problems posed as stories so called algebra word problems are amatter of special difficulty for many students. Use of a computer program tomodel the processes of interpreting such problems, and comparison of the modelwith students' behaviors, reveals that there are (at least) two strikingly differentstyles for tackling these tasks. One style is based on a literal, grammaticaltranslation of the problem statement; the other on conceiving a realworldsituation that corresponds to the problem statement. Most students use one orthe other of these processes almost exclusively and appear largely unaware of the


one they don't use. The ablest students. however, tend to use both processes inreinforcement of each other. This hind of process model can now he developedfurther as a tool to diagnose indivLdual students' difficulties with word problems,and as a framework for developing any additional skills that students are foundto lack.

There are today a substantial number of research groups using computersimulation methods in conjunction with laboratory experiments with humansubjects in order to study complex thought processes, including language com-prehension processes. These research groups can be found both in psychologydepartments and in the "artificial intelligence" divisions of computer sciencedepartments. At M.I.T., for example, a language understanding system has beenconstructed that answers questions and performs actions in a simulated "toyworld" of blocks of various shapes and colors. At the Stanford ResearchInstitute, a group is building a language-understanding "robot," capable oflocomoting about the laboratory rooms. At the University of California, SanDiego, a large "semantic net" simulated on a computer is used to study howhumans store and retrieve the many varieties of information they need for suchactivities as preparing a meal in a kitchen. At Carnegie-Mellon University acomputer has been programed to use phonological, grammatical, and semanticcues to understand spoken language. These are but a sample of the manyresearch activities where the computer is and could be used to explore the natureand processes of human language comprehension.

Concern with understanding cognitive processes is not limited to researcherswho use computer simulation as their main research tool. By timing carefully therelative speeds with which persons can perform various related language compre-hension tasks, cues can be obtained as to the processing that is going on. That aperson can answer the question, "Is John taller than Bob?" a tenth of a secondfaster than "Is Bob shorter than John?" may seem a curious but isolated fact.On the contrary, it is a fact that was predicted by linguistic theory, on the basisthat "taller" is positively associated with the underlying dimension "height,"while "shorter" is negatively associated with that dimension. Tens and hundredsof experiments about many such seemingly innocent features of language arebeginning to tell us what skills a child must acquire in order to understand thespoken or written word.

Language understanding processes are also being illuminated by experimentsof more traditional design. A written passage is a complex "stimulus" for anexperimental subject. What a reader attends to and extracts from this stimulusdepends very much upon instructions and other preconditions that determine hisexpectations and attention when he begins reading. The same material may he"read" and interpreted in a wide variety of ways in response to differentinstructions. Experiments like these begin to give us a scientific basis forclassifying "styles" of reading, and for discovering the conditions that induceone or anothe style to be adopted by the reader.

42 Linguistic Communication: Perspectives jim Research

Additional studies an, needed that will look at understanding with andwithout external memory aids ( paper and pencil, blackboard and chalk), thatwill require people to combine injimnation acquired through reading withinformation presented in pictorial form, and that will compare one's understand-ing of identical information presented orally and in writing. Results of suchexperiments should broaden our understanding of comprehension by extendingit to other ways of encoding the information that students are expected to learn.

Certain technological developments may be ready for application. Forexample, animated film has potential as a teaching aid, but the time required to"shape up" a lesson through several revisions makes animation expensive andslows the development process. Studies of picture-text interaction using stillpictures, and studies of verbal and nonverbal comprehensibility using cinematictechniques are dreadfully expensive for the same reason. On the other hand.there are commercial developments that should help here, and support should begiven to researchers attempting to develop alternative devices. There are com-mzrcial computergraphics concerns and facilities that should make compilingand revising pictorial displays relatively simple', and there are expensive editingmachines that transfer all film to tape, permit computer editing, and thencompile the edited film.

a. Measurement of comprehension and comprehens:hility. Comprehensionis tested by determining whether people, after reading a message, can answerquestions about its contents or implications or can carry out its instructions.Various comprehension tests are in use, but they are in general not derived fromany systematic theory, and they are all heavily dependent on verbal abilities intheir execution. These may be serious limitations. A recurring theme in many ofthe presentations that were heard by this committee was that our verbalmeasures of how well a reader comprehends a written message are inadequateindices of what was grasped. Even more important is the fact that research withrelatively nonverbal measures of comprehension suggests that verbal measures ofan individual's reading ability do not agree with other measures of his ability toextract meaning (e.g., with the ability to comprehend and remember the mean-ing and arrangement of complex pictures).

We need nonverbal measures of comprehension. Such tools should becapable of measuring the comprehension of events that are relevant and familiarto the student (e.g., social relationships), and should not be tied to the items ofthe classroom nor to the articles of a particular (majority) culture. Picturesmight play an important part in the development of such measures. For exam-ple, we might measure a reader's comprehension of a textual passage by itseffects on the speed and accuracy with which the meaning of an otherwiseambiguous picture is apprehended. Alternatively, we might measure the accuracywith which either verbal or nonverbal instructions are followed, using nonverbalbehavior as the test.


Measures of comprehension are needed not only to assess the performanceof individuals or groups of subjects they are also needed for the evaluation ofhow comprehensible a given message is to sonic group of readers. We need suchmeasures of comprehensibility if we are to make the job of reading as easy aspossible. Comprehensibility indices have been applied to textbook designs,advertising and instruction writing, but again the measures used bear littlerelation to any systematic theory of comprehension, and are specifically verbalin nature. Nonverbal measures of comprehensibility and measures of nonverbalcomprehensibility, applicable to pictures and to other methods of communica-tion, would be particularly valuable.

The issue of nonverbal comprehension raises the question of how best to usethe various channels of information that we possess. Some kinds of informationand concepts are obviously better presented by pictures (and especially bymotion pictures and the video media) than by words. The ability to readefficiently depends in part on the reader's knowledge about the world, in ageneral sense, regardless of the channel through which he originally acquired thatknowledge. It seems highly desirable to undertake research directed at determin-ing what kinds of knowledge are better understood, retained and put to use (cf.the "visual thinking" tradition) when presented in pictorial rather than in verbalform. Such research would use measures of verbal and nonverbal comprehensionto determine the principles that govern the comprehensib...ty of pictorial andtextual material, as well as to determine the comprehension that has beenachieved by a particular reader or category of readers.

b. Measures of attention. Similar steps are needed in the development ofattention measures. Particularly important here are questions of unobtrusivenessof the recording devices, closeness to online conditions, and wide validity.Brainwave devices (e.g., central nervous system measures) are probably notusable at the present state of the art. Pupilary diameter appearsto be a measuremore of cognitive load than of attention as we normally want to measure it, andthe history of attempts to measure it automatically is not encouraging.

3. WritingWe have said that reading and writing are mutually supporting skills. Just as

little is known about the stages a person goes through in learning to read, solittle (even less) is known about stages in learning to write from early scribblesto simple but complete and coherent compositions.

A useful way to start would be to make detailed, longitudinal case studies ofindividual children encouraged to write as well as read. Now does the form of achild's writing evolve from isolated letters and words to a faithful representationof his spoken language? What are the stages by which spelling shifts from

44 Linguistic Communication: Perspectives jbr Research

invented solutions toward more standard orthography? In multi-sentence com-positions, what thematic organizations do young children use spatial, temporalor other? What seem to be the relations between stages in learning to write andstages in learning to read? And what are the similarities and differences acrosschildren?

4. Second-language Learning

For children who come from non-English-speaking homes, literacy inEnglish must grow from a foundation of acquisition of oral English as a secondlanguage. Although there is now a considerable body of research on the child'sacquisition of his first language, the same detailed descriptions do not exist forsecond-language acquisition. Here, as elsewhere, we stress the importance ofdetailed analytic studies of learning under varied environmental conditions,rather than comparisons of instructional programs.

How do stages in the acquisition of English as a second language differ fromstages in the acquisition of English as a native language? Do these stages differ innature or rate at different ages? For example, if a Spanish-speaking child istaught in his native language during the beginning years of school and learnsEnglish at 7 or 8 rather than 5 or (), what are the costs and what are the gains?And if that child learns to read first in Spanish, is there positive transfer (ornegative interference) when that child confronts the task of learning to read inEnglish?

Although a variety of bilingual education programs should he supportednow, there will he limits to our attempts to improve education for non-English-speaking children without better answers to these questions.

It should he obvious from this review that the Study Group felt a need forinure research on a wide variety of processes involved in or related to the readingprocess and the learning of reading. This opinion should not he taken to meanthat good research has not already been done the quality of research in thisfield has improved markedly during the past decade as more scientists havebecome interested in it. Indeed, it is this recent revival of interest that encour-aged us to believe that it would he possible to advance rather rapidly toward theresearch goals that we have outlined, and that the NIE would be well advised toplace high priority on exploiting the opportunities that presently exist.


References

Carroll, John B. A model of school learning. leachers College Record, 14e13,64(8).723-733.

Frase, Lawrence T. Maintenance and control in the acquisition of knowledgefrom written materials, In R. Freedle & J. B, Carroll (Eds.i languagecomprehensiem and the acquisition of knowledge. Washington. D. C. V. 11.Winston, 1972. Pp. 337-357.

Jencks. Christopher, et al, Inequality: A reassessment of the effect of family andschooling in America. New York: Basic Books, 1972.

Labov, William. Language, in the Inner City: Studies in the Black EnglishVernacular. Philadelphia: University of Pennsylvania Press, 1972.

Northcutt, Norvell W. Adult performance level study. University of Texas atAustin. 1973,

Schrader, W, B, Test Data as Social Indicators. Educational Testing Service,September 1968.

Terman, L. M. & Merrill. Mark A. Stan lord-Binet Intelligence. Seale: Man fiirthe Third Revision, Form i.M. Boston: Houghton Mifflin, 1960.

Tuddenham, Read D. Soldier intelligence in World Wars I and II. AmericanPsychologist. 1948.3.54-56.

d c - eric · cs 001 353. miller, george a., ed. linguistic communication: perspectives for...

Documents